Form plugin remote code execution vulnerability blocked more than 400,000 times, should the form be withdrawn from the built-in website?
The remote code execution vulnerability of the form plugin has been blocked more than 400,000 times. Should we withdraw the form to the in-built station - it is worth recalculating this account from the perspective of the attack surface.The statistics disclosed by the security agency show that the form plugin Super Forms in the WordPress ecosystem had a high-risk vulnerability (CVE-2024-14894) before version 6.3.314, with a rating of 9.8, and it is a level where remote code execution can be performed as soon as it is installed;Page builder Elementor Pro also has a CVE-2024-32475 number and a 7.5 exploit risk rating.The protective product intercepted over 440,000 exploitation attempts in a single statistical period. The value of the number lies in: it is not a theoretical threat model, but a real queuing attack volume.
Why is the entry-type function being attacked
Forms, comments, login, and other such features are among the few places on a site that allow strangers to write in. Plugin solutions usually stack capabilities for universality: file upload, rich text, and triggering third-party actions, each layer of capability is a new parser, and thus a new candidate for vulnerabilities. A rating of 9.8 means it can be triggered without an account - attackers do not need to register first, and scanners can bulk deliver to public URLs in the hope of hitting the jackpot. Behind the more than 400,000 attempts, the plugin market has standardized 'entry points', and the attack script has also been standardized.
The reality of patching the plugin layer
The handling chain of plugin vulnerabilities is longer than that of the core program: manufacturers release announcements, authors release patches, update channels distribute, and sites complete upgrades. There are often存量plugins in the ecosystem that are still in use despite the authors stopping updates, and these will never be automatically updated. Choosing plugins is not impossible, but it should be able to answer three questions: does this plugin have security update records in the past half year, how fast is the response speed when problems occur, and is there an alternative solution that can take it offline.
To withdraw the built-in, you need to consider which three capabilities
The form and message withdrawal system is built-in, essentially replacing the 'third-party parser' with 'audited code in the main program'. Taking AnQiCMS as an example, after withdrawal, three layers of protection must be fully opened: the first layer is submission verification, the form is connected to reCAPTCHA verification code, blocking automated delivery before parsing, and no matter how large the attack volume is, it is only rejected robot requests.The second layer is the content layer, with preliminary processing of sensitive word filtering and keyword replacement, accompanied by manual confirmation in the queue for content review; the third layer is the foundation defense, with built-in JWT authentication and protective logic against SQL injection and XSS attacks, and the parsing layer itself does not leave any loopholes for transparent transmission and concatenation.All three layers are within the scope of the security audit of the main program version, adding one version is equivalent to covering all three layers of the sites, and there is no long tail of 'some plugin has not been updated yet'.
What scenarios should still be left at the plugin level
Need a complex approval process, connect with multiple CRM systems, and deeply customized form interactions, modules that come with it may not cover everything, this kind of thing should stay on the plugin layer, but you need to set discipline: only install actively maintained ones, apply security updates in a timely manner, add verification and rate limits to the form endpoints separately, and regularly audit components that have been installed but not used. Security selection is never a choice between two, but determines which part of the attack surface is protected by your own version release rhythm.
Common questions
Question: Don't the built-in forms have any vulnerabilities?Answer: Not at all, but convergence. The attack surface has shrunk from 'main program plus dozens of plugin parsers' to 'a single audited entry', and fixes are released with the main program in one go, covering all.
Question: Can reCAPTCHA replace the review?Answer: No. It blocks robot submissions, but cannot block spam and违规 content sent by real people; captcha, sensitive word filtering, and manual review each handle a section.
Question: Is the Anqi CMS comment function suitable for directly incorporating on-site form requirements?Answer: Light collection (consultation, comments, registration) is enough, with captcha and approval queue you can go online; complex process arrangement still requires professional form tools, keep the plugin layer but add the discipline above.